Summary 1. Objective 2.1: Investigation of novel pathways regulating HSC self-renewal and differentiation By exploiting the extensive HSC amplification after transplantation, we aim to identify novel factors/pathways involved in HSC self-renewal. We have performed experiments to identify conditions for optimal HSC amplification after transplantation, including cell dose, timing of cell collection after transplantation, and the impact of serial transplantations. We found that lower cell doses (e.g. 1x106 cells) result in superior HSC expansion in vivo compared to larger cell doses (e.g. 1x107 cells). The optimal timing for cell collection was 1 week; longer times in vivo (e.g. 2, 3, or 4 weeks) resulted in progressive differentiation to more mature progeny. Serial transplantations have led to engraftment levels too low to allow reliable detection and selection of HSCs. In FY17, we will compare global transcriptome and methylation patterns between HSC at steady state (before transplant) and after transplant, using RNA-Seq and CHIP-Seq approaches. A number of transcription factors involved in the maintenance of HSC function have been identified and recent studies have suggested involvement of epigenetic mechanisms to orchestrate the activities of these factors to ensure blood homeostasis. DNA methylation of CpG dinucleotides is a key epigenetic modification that influences mammalian gene expression. CpG methylation is catalyzed by a family of DNA methyltransferase (DNMT) enzymes comprising three members, DNMT1, DNMT3a, and DNMT3b. Conditional ablation of DNMT3a in a mouse model resulted in progressive impairment of HSC differentiation over serial transplantation, while simultaneously expanding HSC numbers in the bone marrow. DNMT3a-null HSCs upregulated HSC multipotency genes and downregulated differentiation factors, and their progeny exhibited global hypomethylation and incomplete repression of HSC-specific genes. These data established DNMT3a as a critical participant in the epigenetic silencing of HSC multipotency genes, thereby enabling efficient differentiation. In FY17, we intend to extend these findings by investigating whether transient downregulation of DNMT3a expression by CRISPR interference in human CD34+ cells may prevent differentiation of HSCs and favor their self-renewal during short-term expansion ex vivo. 2. Objective 2.2: Potentiation of known and new pathways involved in HSC self-renewal with hypoxia. Activation of Notch signaling in human HSPCs by treatment with Notch ligand Delta1 has enabled a clinically relevant ex vivo expansion of short-term HSPCs. In vitro studies have also revealed a role of low O2 tension in HSPC regulation. A molecular link has been demonstrated in several stem/progenitor cell populations between Notch and hypoxia pathways but their interaction has not been investigated in human HSPCs. G-CSF mobilized human CD34+ cells from 4 healthy subjects were cultured in the presence of cytokines (SCF, FLT3L and TPO) in hypoxia (1.5-2% O2) or normoxia (21% O2) in vessels coated with fibronectin alone or combined with increasing concentrations of the immobilized ligand Delta1 (2.5, 5, 10 and 20 g/mL). After 21 days in culture, cells were counted and characterized using CFU assays, flow cytometry for lineage (Glycophorin A+, CD13+, CD20+, CD3+ and CD41+ cells) and HSC (CD34+ CD38- CD45RA- CD90+ CD49f+ Rholow) phenotypes, and transplantation in immunodeficient (NSG) mice. In normoxia, the total number of cells increased 118-fold compared to baseline in the absence of Delta1 with limited residual CD34+ cells (1.5 0.7%), extensive differentiation toward the myeloid lineage (96.3 0.3% CD13+ cells) and minimal engraftment potential in NSG mice (0.2 0.2% human CD45+ cells). With increasing concentrations of Delta1 in normoxia, consistent with the hypothesis that Delta1 delays differentiation, the total number of cells increased less (41-, 25-, 11- and 7-fold relative to baseline, respectively) CD34+ cells expanded more (4-, 4-, 3- and 2-fold relative to baseline, respectively), and CFU numbers increased more (8-, 7-, 4- and 3-fold relative to baseline, respectively) than without Delta1. However, phenotypically defined HSCs were undetectable or markedly decreased at the lowest Delta1 concentrations used (2.5 and 5 g/mL) and their numbers were maintained or only minimally increased at the highest Delta-1 concentrations tested (10 and 20 g/mL) relative to uncultured CD34+ cells. Accordingly, only cells cultured with 10 and 20 g/mL Delta1 resulted in levels of engraftment in NSG mice (5.5 5.4% and 5.4 0.9% human CD45+ cells, respectively) comparable to uncultured cells (7.0 0.1% human CD45+ cells). In hypoxia, total cell counts increased less than in normoxia both without (8-fold relative to baseline) and with increasing concentrations of Delta1 (11-, 11-, 9-, 9-fold relative to baseline, respectively) due to diminished myeloid differentiation. Total CD34+ cells decreased 1.7-fold in hypoxia in the absence of Delta1, but expanded modestly in the presence of Delta1 (3-, 3-, 2- and 2-fold, respectively). CFU numbers followed a similar trend. However, in hypoxic cultures with 2.5, 5 and 10 g/mL Delta1, phenotypically defined HSCs increased 2.5-, 6.6- and 1.3-fold, respectively, compared to uncultured cells. Importantly, hypoxia combined with 2.5, 5 and 10 g/mL Delta1 concentrations resulted in increased human cell engraftment in NSG mice (21.2 4.4%, 29.3 11% and 11.8 5.4% human CD45+ cells, respectively) compared to uncultured cells (7.0 0.1% human CD45+ cells). When 20 g/mL Delta1 was used in hypoxia, engraftment potential in NSG mice was decreased (1.1 0.6% human CD45+ cells). We next performed limiting dilution analysis to measure the frequencies of long-term repopulating HSCs (LT-HSCs) within the CD34+ cell compartment at baseline and after 21 days in hypoxic or normoxic cultures supplemented with the optimized concentrations of Delta1 (10 g/mL in normoxia and 5 g/mL in hypoxia). LT-HSCs in uncultured CD34+ cells were measured at the expected frequency (1 in 7,706; 95% CI of 3,446 to 17,232). When analyzed at 3 months post-transplantation, a limited (1.5-fold) increase in LT-HSC frequency (1 in 5,090; 95% CI 2.456 to 10,550) was obtained from Delta1 normoxic cultures compared to uncultured cells. In contrast, the frequency of LT-HSCs (1 in 1,586; 95% CI 680 to 3,701) was 4.9-fold higher in hypoxic Delta1 cultures compared to uncultured cells, and 4.2-fold higher than in normoxic Delta1 cultures. Similarly, absolute numbers of LT-HSCs per 100,000 Day 0 equivalent CD34+ cells increased from 13 (baseline) to 216 (normoxia) and 694 (hypoxia). Our data indicate that hypoxia potentiates Notch-induced expansion of human HSPCs and may be of benefit in stem cell transplantation and gene therapy applications. In FY 17, we will confirm these findings in a large animal transplantation model (rhesus macaques) that permit long-term evaluation after transplantation. We will identify the molecular mechanisms underlying the intersection between Notch and hypoxia pathways.